[Mlir-commits] [mlir] [mlir][python] implement GenericOp bindings (PR #124496)

[Maksim Levental]

https://github.com/makslevental created https://github.com/llvm/llvm-project/pull/124496

None

>From 206cb67df44b6dade7408a4adb8337ca01e00c82 Mon Sep 17 00:00:00 2001
From: max <maksim.levental at gmail.com>
Date: Sun, 26 Jan 2025 20:01:51 -0500
Subject: [PATCH] [mlir][python] implement GenericOp bindings

---
 mlir/python/mlir/dialects/linalg/__init__.py | 44 ++++++++++++++
 mlir/test/python/dialects/linalg/ops.py      | 60 ++++++++++++++++++++
 2 files changed, 104 insertions(+)

diff --git a/mlir/python/mlir/dialects/linalg/__init__.py b/mlir/python/mlir/dialects/linalg/__init__.py
index 8fb1227ee80ff5..946094e2e9f691 100644
--- a/mlir/python/mlir/dialects/linalg/__init__.py
+++ b/mlir/python/mlir/dialects/linalg/__init__.py
@@ -10,6 +10,7 @@
 #   DSL -> YAML -> tblgen -> pytblgen -> build/.../_linalg_ops_gen.py.
 from .._linalg_ops_gen import *
 from .._linalg_enum_gen import *
+from .._linalg_enum_gen import _iteratortypeenum
 
 # These are the ground truth functions defined as:
 # ```
@@ -58,6 +59,7 @@
 
 from ...ir import *
 from .._ods_common import get_op_result_or_value as _get_op_result_or_value
+from ...extras.meta import region_op
 
 
 def transpose(
@@ -102,3 +104,45 @@ def broadcast(
     )
     fill_builtin_region(op.operation)
     return op
+
+
+ at register_attribute_builder("IteratorTypeArrayAttr")
+def _IteratorTypeArrayAttr(x, context):
+    return ArrayAttr.get([_iteratortypeenum(v, context) for v in x])
+
+
+class GenericOp(GenericOp):
+    def __init__(
+        self,
+        inputs,
+        outputs,
+        indexing_maps,
+        iterator_types,
+        *,
+        doc=None,
+        library_call=None,
+        loc=None,
+        ip=None,
+    ):
+        result_types = []
+        if isinstance(outputs[0].type, RankedTensorType):
+            result_types = [o.type for o in outputs]
+
+        super().__init__(
+            result_types,
+            inputs,
+            outputs,
+            indexing_maps,
+            iterator_types,
+            doc=doc,
+            library_call=library_call,
+            loc=loc,
+            ip=ip,
+        )
+        element_types = [i.type.element_type for i in inputs] + [
+            o.type.element_type for o in outputs
+        ]
+        self.regions[0].blocks.append(*element_types)
+
+
+generic = region_op(GenericOp, terminator=YieldOp)
diff --git a/mlir/test/python/dialects/linalg/ops.py b/mlir/test/python/dialects/linalg/ops.py
index 72045a07b2da80..b7e0f2884bb249 100644
--- a/mlir/test/python/dialects/linalg/ops.py
+++ b/mlir/test/python/dialects/linalg/ops.py
@@ -84,6 +84,7 @@ def named_form(lhs, rhs):
 
     print(module)
 
+
 # CHECK-LABEL: TEST: testIdentityRegionOps
 @run
 def testIdentityRegionOps():
@@ -161,3 +162,62 @@ def broadcast_op(op1, op2, op3):
                 op5 = linalg.broadcast(op3, outs=[op2], dimensions=[0])
 
     print(module)
+
+
+# CHECK-LABEL: TEST: testGenericOp
+ at run
+def testGenericOp():
+    with Context(), Location.unknown():
+        module = Module.create()
+        f32 = F32Type.get()
+        with InsertionPoint(module.body):
+            id_map = AffineMap.get_identity(2)
+            # CHECK: %[[VAL_0:.*]] = tensor.empty() : tensor<16x16xf32>
+            # CHECK: %[[VAL_1:.*]] = tensor.empty() : tensor<16x16xf32>
+            x = tensor.empty((16, 16), f32)
+            y = tensor.empty((16, 16), f32)
+
+            # CHECK: %[[VAL_3:*]] = linalg.generic {indexing_maps = [#map, #map], iterator_types = ["parallel", "parallel"]} ins(%[[VAL_0]] : tensor<16x16xf32>) outs(%[[VAL_1]] : tensor<16x16xf32>) {
+            # CHECK: ^bb0(%in: f32, %out: f32):
+            # CHECK:   linalg.yield %in : f32
+            # CHECK: } -> tensor<16x16xf32>
+            @linalg.generic(
+                [x],
+                [y],
+                [id_map, id_map],
+                [linalg.IteratorType.parallel, linalg.IteratorType.parallel],
+            )
+            def f(x, y):
+                return x
+
+            assert isinstance(f, Value)
+
+            # CHECK: %[[VAL_3:.*]] = tensor.empty() : tensor<16x16x16xf32>
+            z = tensor.empty((16, 16, 16), f32)
+
+            minor_id = AffineMap.get_minor_identity(3, 2)
+            id_map = AffineMap.get_identity(3)
+
+            # CHECK: %%[[VAL_4:.*]]:2 = linalg.generic {indexing_maps = [#map1, #map2, #map2], iterator_types = ["parallel", "parallel", "parallel"]} ins(%[[VAL_0]] : tensor<16x16xf32>) outs(%[[VAL_3]], %[[VAL_3]] : tensor<16x16x16xf32>, tensor<16x16x16xf32>) {
+            # CHECK: ^bb0(%in: f32, %out: f32, %out_0: f32):
+            # CHECK:   linalg.yield %in, %out : f32, f32
+            # CHECK: } -> (tensor<16x16x16xf32>, tensor<16x16x16xf32>)
+            @linalg.generic(
+                [x],
+                [z, z],
+                [minor_id, id_map, id_map],
+                [
+                    linalg.IteratorType.parallel,
+                    linalg.IteratorType.parallel,
+                    linalg.IteratorType.parallel,
+                ],
+            )
+            def g(x, z1, z2):
+                return x, z1
+
+            assert isinstance(g, OpResultList)
+            assert len(g) == 2
+            assert isinstance(g[0].type, RankedTensorType)
+            assert isinstance(g[1].type, RankedTensorType)
+
+    print(module)